Handwriting Input Candidate String Storage for Recognition Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current handwriting input systems on touch screens face challenges in miniaturization, navigation, and editing of text entries, particularly on smaller devices like smartphones, where traditional methods are inadequate for converting graphic objects into digital symbols effectively.
Innovation Solution
A touch screen device with an interface unit and a graphic object processing module that detects and records graphic data, generates candidate strings with relevance values, and allows users to select and insert strings into an input field, with the ability to store and modify data for subsequent use in navigation and editing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Volume of moving object
If handwriting input systems are adapted for miniaturized touch devices, then device portability and accessibility are improved, but recognition accuracy and input precision deteriorate
Solution Approach 1:
The system segments the handwriting recognition process into multiple stages: initial candidate string generation, relevance value assignment, and subsequent selection or correction. This multi-stage approach allows accurate recognition even on miniaturized devices by processing graphic data systematically rather than requiring perfect single-pass recognition.
Solution Approach 2:
The system performs preliminary action by generating multiple candidate strings with relevance values before final insertion. This preliminary processing maintains recognition accuracy on small devices by providing a buffer of possible interpretations, allowing the most relevant candidate to be selected even when input precision is limited.
2Device complexity
If traditional handwriting conversion methods are used, then system simplicity is maintained, but navigation and editing functionalities are insufficient
Solution Approach 1:
The system implements multi-functionality by integrating navigation and editing capabilities into the handwriting input process. The same graphic data and candidate strings used for initial input are reused for navigation and correction, allowing a single system structure to handle multiple functions without proportionally increasing complexity.
Solution Approach 2:
The system uses candidate strings with relevance values as an intermediary between the handwritten input and the final text output. This intermediary structure enables both navigation (by selecting different candidates) and editing (by correcting selected strings), enhancing ease of operation without requiring a completely separate complex editing system.
3Speed
If graphic data is stored temporarily for recognition, then processing speed is improved, but data association and editing capability are lost
Solution Approach 1:
The system performs preliminary action by generating and storing multiple candidate strings with their associated relevance values alongside the graphic data. This preliminary storage of associated information maintains the link between graphic input and possible interpretations, enabling later editing while not significantly impacting initial processing speed.
Solution Approach 2:
The system ensures continuity of useful action by maintaining the association between graphic data and candidate strings throughout the process. This continuous preservation of data relationships allows the system to transition smoothly from rapid initial recognition to subsequent editing operations without losing critical information associations.
Data Source
Figure 1A~2
Figure 3A~3D
Figure 3E~4B
AI summary
The invention proposes a device (2) for inputting symbols into an input field (30), said device comprising an interface unit comprising a touch screen (18) and a graphical object processing module, said module comprising: • a unit for detecting a start of input and an end of input of a current graphical object (OG1); • a recording unit for recording graphical data corresponding to the input graphical object (OG1); • a recognition unit for generating a list (LC1) of candidate strings (C11-C14) of symbols from the graphical data, each candidate string (C11-C14) being associated with a relevance value; and • an insertion unit for inserting, into the input field (30), a string (CS1) selected by the user from the list, a group of data items comprising the graphical data, the candidate strings (C11-C14), the relevance values, and an identifier of the selected string (CS1), being stored for a predefined period.